A variety of studies has attempted to define etiologic factors for cancers of the breast, endometrium, and prostate. In a study of breast cancer among younger women, induced abortion was not found to be related to risk, nor were a variety of past medical conditions. In this same study, risk factors for in situ cancers generally appeared to be similar to those identified for invasive disease. A study of breast cancer in Asian-Americans showed that menstrual and reproductive factors appeared to have similar relationships as in higher-risk Caucasian populations. To assess the effects on breast cancer risk of different types and indications for gynecologic surgery, a record linkage study was undertaken in Sweden, showing substantial reductions in risk for women oophorectomized at young ages, but little alteration in risk for those undergoing a simple hysterectomy. Incidence data from the Surveillance, Epidemiology and End Results Program (SEER) confirmed an overall 5% excess of left-sided breast cancer, possibly reflecting that this side of the breast tends to be larger than the right side. In a series of methodologic studies, the reproducibility of various endogenous hormonal markers was assessed. Those assays shown to be reproducible were assessed in relation to endometrial cancer, showing increased risk associated with high levels of androstenedione among both premenopausal and postmenopausal women. High levels of estrone and bioavailable estradiol were related to increased disease risk among postmenopausal women, and appeared to be independent of the effects of obesity or other risk factors. Ongoing studies will assess endogenous hormones in relation to breast cancer from both the studies of breast cancer in younger women and in Asian-Americans. A recently completed case-control study of prostate cancer in Shanghai, China will attempt to assess a variety of hormonal factors in relation to risk. Assessment of hormonal factors in relation to colorectal cancer failed to support a link with menstrual or reproductive factors.